Available online at www.sciencedirect.

com
Journal of
CONSUMER
PSYCHOLOGY
Journal of Consumer Psychology 18 (2008) 116 – 126

Consumer response to brand extensions: Construal level as a moderator of

the importance of perceived fit
Hakkyun Kim a,1,2 , Deborah Roedder John b,⁎,2
a
John Molson School of Business, Concordia University, Montreal, QC, Canada H3G 1M8
b
Carlson School of Management, University of Minnesota, Minneapolis, MN 55455, USA
Received 20 January 2008
Available online 14 March 2008

Abstract

One of the most enduring findings from branding research is that consumers evaluate brand extensions on the basis of their perceived fit with
the parent brand. In this article, we propose that the importance of perceived fit in extension evaluations is moderated by construal level. We draw
upon construal level theory, which posits that individuals can construe stimuli in their environments in terms of abstract and generalized features
(high-level construals) or in terms of concrete and contextualized features (low-level construals). Results from three studies confirm that
consumers who construe their environment at a higher level place more importance on perceived extension fit in evaluating brand extensions.
These consumers evaluate high fit extensions more favorably than moderate fit extensions, consistent with prior research. However, consumers
who construe their environment at a lower level do not evaluate high and moderate fit extensions any differently, unless the importance of using fit
perceptions is made salient.
© 2008 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved.

Brand extensions are one of the most popular strategies for et al., 1987). Perceived fit can be based on the extension being
leveraging brand equity. By launching new products under in a product category similar to other products sold by the
popular brand names, firms hope that consumers will respond parent brand (Boush et al., 1987; Keller & Aaker, 1992),
more favorably to the new offering, due to their familiarity with complementing use with other products sold by the parent
the parent brand, positive feelings toward the parent brand, and brand (Aaker & Keller, 1990), being in a product category
positive attribute and non-attribute associations they have with where the parent brand can contribute an appealing attribute
the parent brand. Compared to launching a new product under a (Broniarczyk & Alba, 1994; Herr, Farquar, & Fazio, 1996),
new brand name, brand extensions can increase the efficiency of having a parent brand with the skill and expertise to make the
promotional efforts, improve access to distribution channels, extension product (Aaker & Keller, 1990), and having a parent
and reduce consumers' perceived risk of purchasing a product brand with an image that is compatible with the extension
or service (Keller, 2002). (Park, Milberg, & Lawson, 1991). Perceived fit can also be
What factors determine whether or not a brand extension heightened by communications providing a plausible link
will be successful? The most important factor identified by between the extension and parent brand (Bridges, Keller, &
prior research is perceived fit. Consumers respond more Sood, 2000), thereby giving consumers more opportunity to
favorably if they are able to perceive a fit between the discover possible links (Lane, 2000), or countering negative
extension and the parent brand (Aaker & Keller, 1990; Boush inferences about the extension (Aaker & Keller, 1990).
Perceived fit, no matter how it is defined, is the most important
determinant of brand extension success—more important than
marketing support, retailer acceptance, and quality of the
⁎ Corresponding author. Fax: +1 612 626 8328.
parent brand (Völckner & Sattler, 2006).
E-mail addresses: [email protected] (H. Kim), [email protected]
(D.R. John). In fact, the role of perceived fit is so widely accepted that
1
Fax: +1 514 848 4554. researchers have largely neglected the possibility that moderat-
2
Both authors contributed equally to this manuscript. ing factors may exist that render perceived fit more or less
1057-7408/$ - see front matter © 2008 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved.
doi:10.1016/j.jcps.2008.01.006
 H. Kim, D.R. John / Journal of Consumer Psychology 18 (2008) 116–126 117

influential in brand extension evaluations. The working making purchase decisions (Dhar & Kim, 2007; Kardes,
assumption has been that, for all types of consumers, perceived Cronley, & Kim, 2006; Trope, Liberman, & Wakslak, 2007).
fit is always important for evaluating brand extensions. To date, We add to this line of inquiry by examining the role of construal
the only evidence of moderating factors is found in cross- level in consumer response to branding strategies, specifically
cultural settings, where researchers have reported that con- brand extensions. Our empirical findings are in line with the key
sumers from Eastern cultures place less importance on conceptual premise of construal level theory and respond to a
perceived fit in evaluating brand extensions, relying instead recent call for more research on the role of construal level in
more on background factors, such as corporate brand reputation consumer behavior contexts (Liberman, Trope, & Wakslak,
(Bottomley & Holden, 2001; Han & Schmitt, 1997). 2007; Lynch & Zauberman, 2007).
In this paper, we propose that construal level is a key
moderator of the importance of perceived fit in brand extension Conceptual background
evaluations. We draw upon construal level theory, which posits
that individuals can construe stimuli in their environments in Construal level
different ways—in terms of abstract and generalized features
(high-level construals) or in terms of concrete and contextua- Construal level theory (CLT) posits that individuals
lized features (low-level construals) (Trope & Liberman, 2003). construct different representations of stimuli in their environ-
In judgment and decision making settings, individuals favor ments, which vary in terms of the degree of abstraction (Freitas,
information, experiences, or events that match their construal Salovey, & Liberman, 2001; Trope & Liberman, 2003;
level (Nussbaum, Trope, & Liberman, 2003; Trope & Liber- Vallacher & Wegner, 1987). Individuals who use more abstract
man, 2000). Thus, we reason that consumers who construe mental models construe stimuli with relatively simple, decon-
stimuli in terms of abstract and generalized features (high-level textualized, and coherent representations that extract the gist
construals) will place more importance on perceived fit, which from available information (high-level construals). Therefore,
is a global and abstract assessment of the connections that exist these people tend to be influenced by abstract and general
between an extension and parent brand, when evaluating brand features of stimuli, such as stereotypical features that are the
extensions. In contrast, consumers who construe stimuli in result of abstraction and generalization about the characteristics,
terms of concrete and contextual features (low-level construals) attributes, and behaviors of certain types of people, events, or
will place less importance on perceived fit and more importance other stimuli (Ashmore & Del Boca, 1981; Fiske & Pavelchak,
on more concrete features of the brand extension when 1986; Hilton & von Hippel, 1996). In contrast, people who use
evaluating extensions. concrete mental models construe stimuli with relatively
We examine this proposition in three experiments. In study 1, complex, contextualized, and incidental representations (low-
we examine construal level as an individual difference variable. level construals). Consequently, these individuals tend to be
We compare consumers with a chronic tendency to construe influenced by specific and detailed features of stimuli, as well as
their environment in an abstract way (high-level construals) contextual details salient at the moment (Fiske & Pavelchak,
with consumers having a chronic tendency to construe their 1986).
environment in a more concrete way (low-level construals). In Action identification theory supports a similar distinction.
study 2, we examine construal level as a contextual variable. According to this theory, the identities of an action can be
Here, we manipulate construal level by varying temporal arranged in a cognitive hierarchy, from low-level identities
distance, comparing brand extension evaluations taking place in specifying how one acts, to high-level identities specifying why
the near future (low-level construals) versus the distant future one acts (Vallacher & Wegner, 1989). For example, for those
(high-level construals). Across both studies, we find that thinking about “eating breakfast” with an abstract mindset
construal level moderates the importance of perceived fit, (high-level construals), “getting nutrition for the day” will make
with greater importance being placed on perceived fit in more sense to them than will “pouring milk on cereal.” For
evaluating brand extensions when high-level construals are those thinking about “eating breakfast” with a concrete mindset
favored by consumers. Moreover, in study 3, we examine a (low-level construals), “pouring syrup on a stack of pancakes”
strategy for encouraging consumers who favor low-level will resonate more with them than will “boosting energy for the
construals to use perceived fit as important input into brand day.”
extension evaluations. Providing a cue for the importance of Construal levels can be determined by situational or
perceived fit, we find that low-construal consumers respond to individual factors. Tests of construal level theory have shown
brand extensions as if they were high-construal consumers. That that psychological distance is a primary determinant of
is, low-construal consumers evaluated high fit extensions more construal level. Individuals tend to use concrete mental models
favorably than moderate fit extensions, consistent with evalua- (low-level construals) when thinking about events in the near
tions of high-construal consumers. future and more abstract mental models (high-level construals)
These findings add to the emergence of construal level when thinking about events in the distant future (Trope &
theory as an important area of consumer behavior research. Very Liberman, 2000). Construal levels can also vary at the
recently, researchers have linked construal levels to important individual level, with individuals having a chronic tendency
aspects of consumer decision making, including the formation toward different levels of construal (Freitas et al., 2001;
of consideration sets, methods for evaluating brands, and Vallacher & Wegner, 1989). As Vallacher and Wegner (1989)
118 H. Kim, D.R. John / Journal of Consumer Psychology 18 (2008) 116–126

note, “…at one extreme is the low-level agent, someone who affectively-based (Edwards, 1990; Fabrigar & Petty, 1999),
operates on the world primarily at the level of details. This strong arguments are more effective when they match the
person tends to approach an action with its mechanistic functional bases of an individual's attitude (Petty & Wegener,
components in mind. At the other extreme is the high-level 1998), and messages emphasizing promotion (vs. prevention)
agent, someone who routinely views his or her action in terms benefits are more persuasive for individuals with goals that are
of causal effects, social meanings, and self-descriptive implica- more promotion focused (Aaker & Lee, 2001). Information that
tions” (p. 661). matches an individual's attitudes, goals, or processing style is
thought to be more influential because it is perceived as more
Influence of construal level on judgment and decision making valid (Lavine & Snyder, 1996), is more important (Lee, Aaker,
& Gardner, 2000), and/or receives more elaboration (Petty &
A considerable body of research demonstrates that construal Wegener, 1998).
levels influence judgments and choices. Studies of psycholo-
gical distance as a determinant of construal level have found Influence of construal level on brand extension evaluations
that differences in construal level lead to differences in the ways
that individuals make judgments and decisions about psycho- Perceived fit has been identified as the most important driver
logically distant versus near events (Fujita, Henderson, Eng, of brand extension evaluation (Völckner & Sattler, 2006).
Trope, & Liberman, 2006; Kardes et al., 2006; Trope & Perceived fit is a global assessment of the connections that exist
Liberman, 2000). As temporal distance increases, preferences between the extension and parent brand, based on factors such
are more likely to be based on the value associated with high- as being in similar product categories, sharing an important
level construals of options than on the value associated with attribute, or complementing each other in usage situations
low-level construals of options. For instance, Trope and (Keller, 2002). Extensions that are perceived as a good fit with a
Liberman (2000) tested a change in the weights of abstract well-regarded parent brand receive favorable evaluations.
versus concrete features of events, according to temporal Extensions that are perceived as a poor fit with a parent brand
distance, by varying the valence of the two types of construals. are evaluated in a less favorable light.
Based on the results of a pretest that “eating a cake” or “a We propose that the importance of perceived fit in brand
mother's party” had a more positive low-level construal than a extension evaluation is moderated by construal level. Specifi-
high-level one, they found that participants rated these cally, we argue that a global and abstract concept, such as
experiences as more positive at a near versus distant future perceived fit, is a better match for consumers with an abstract
time. These results indicate that as temporal distance increases, mindset (who favor high-level construals), with the result that
high-level features become more important than low-level ones perceived fit is more important for these consumers versus those
in judging preferences. with a more concrete mindset. Prior research has shown that
Moreover, construal levels can influence the preference for people with an abstract mindset base their judgments and
certain types of information in making predictions. Nussbaum predictions on higher-level construals of available information,
et al. (2003) report that when participants were allowed to seek whereas people with a concrete mindset base their judgments
new information before making a prediction about a target and predictions on lower-level construals of available informa-
person, they tended to prefer information about more global tion (Liberman & Trope, 1998; Trope & Liberman, 2003).
traits (which are high-level) when the goal was to predict the Because perceived fit is a global judgment based on available
target person's behavior in distant rather than near future information about the extension and parent brand, and global
situations. Confidence in predictions of future events is also judgments of this sort are high-level construals, consumers with
greater for high-level construals. A recent study examining how an abstract mindset are likely to place greater importance on
people use information for prediction finds that people are more perceived fit. However, consumers with a concrete mindset are
confident in theory-based predictions of psychological experi- unlikely to do so because perceived fit is a global judgment that
ments (which are high-level) when the experiments are does not match their propensity to construe stimuli at a more
expected to take place in the distant versus near future concrete and contextualized level. For these consumers, brand
(Nussbaum, Liberman, & Trope, 2006). These findings suggest extension evaluations are more likely to be based on low-level
that individuals with abstract mindsets prefer making predic- construals of available information, such as assessments of the
tions on the basis of information and inferences at higher concrete attributes of the extension product and the purchase/
construal levels. usage context.
In sum, the evidence suggests that construal levels influence In study 1, we examine this proposition by investigating the
judgment and decision making by a preference for information, moderating effect of construal level as an individual difference
experiences, or events that match the individual's abstract or variable. We compare brand extension evaluations for con-
concrete mindset. These findings are reminiscent of the sumers who have a chronic tendency to construe stimuli at a
matching effect found in studies of attitudes and persuasion. global and abstract level (high-level construals) with consumers
Researchers have found that persuasive appeals are more who have a chronic tendency to construe stimuli at a more
effective when they match an individual's goals, attitude bases, concrete and contextualized level (low-level construals). To
or processing styles. For example, emotional appeals are more detect the importance placed on extension fit, consumers were
effective than rational appeals when an individual's attitude is asked to evaluate brand extensions that varied in terms of fit
 H. Kim, D.R. John / Journal of Consumer Psychology 18 (2008) 116–126 119

(high, moderate) with the parent brand. We predict that MNew Balance = 5.38 on a 1 = unfavorable to 7 = favorable scale).
consumers with a chronic tendency toward high-level con- These brands were chosen to be as similar as possible to one
struals will place more weight on perceived extension fit than another, with the exception of Nike being a broader brand than
will consumers with a chronic tendency toward low-level New Balance (MNike = 4.90 vs. MNew Balance = 3.68; p b .01). The
construals. Formally, consideration of both a broad and a narrow brand was deemed
important, as prior brand extension research has shown that
H1. Consumers with a chronic tendency toward high-level
brand breadth can affect brand extension evaluation processes
construals will place more weight on perceived extension fit in
(Boush & Loken, 1991).
evaluating brand extensions compared to consumers with a
To allow for more direct comparisons between brands, we
chronic tendency toward low-level construals.
identified brand extensions that could be used for both brands.
Further, we predict that differences between high fit and Based on a pretest with participants similar to those in the main
moderate fit extensions will be greater for consumers with a study, we selected comfort insoles as the high-fit extension for
chronic tendency toward high-level construals. Prior brand both brands (perceived fit: MNike = 5.86 vs. MNew Balance = 5.83)
extension research has shown that high fit extensions receive and treadmills as the moderate-fit extension for both brands
more favorable evaluations than do moderate fit extensions. For (perceived fit: MNike = 3.73 vs. MNew Balance = 3.38), with
high fit extensions, consumers base evaluations on their positive perceived fit to the Nike/New Balance brand measured on two
beliefs and feelings toward the parent brand, resulting in very 7-point scales (1 = inconsistent/atypical to 7 = consistent/typical).
favorable extension evaluations; however, moderate fit exten-
sions do not benefit from this transfer of positive affect and Sample and procedure
beliefs. Note that the difference between evaluations for high
versus moderate fit extensions assumes that the consumer One-hundred-fifty-eight undergraduates participated in the
places importance on the perceived extension fit. Thus, we study in order to fulfill a course requirement. To begin,
argue that the difference between high and moderate fit participants were asked about their familiarity (1 = unfamiliar
extension evaluations will be greater for consumers who place to 7 = familiar scale) and attitudes toward the brands (1 = unfa-
more weight on the perceived fit (consumers who favor high- vorable to 7 = favorable scale). Next, respondents were shown a
level construals) than others (consumers who favor low-level brief description of one of the brand extensions from either Nike
construals). Formally, or New Balance and were asked for their evaluations of the
extension on three 7-point scales (poor/excellent, inferior/
H2. Differences in brand extension evaluations for high versus
superior, and undesirable/desirable). After providing their
moderate fit extensions will be greater for consumers with a
evaluations, participants were asked to turn to the next page of
chronic tendency toward high-level versus low-level construals.
the survey, where they were asked to list thoughts that came to
mind as they rendered their evaluations. The participants then
Study 1 filled out the Behavior Identification Form, which was used to
identify individuals with a chronic tendency toward low- versus
Experimental design high-level construals. Next, they judged the fit of the brand
extension with the brand (Nike or New Balance) on two 7-point
Our predictions were tested in a 2 (Construal Level: High, scales (inconsistent/consistent and atypical/typical), similar to
Low) × 2 (Brand: Nike, New Balance) × 2 (Extension Fit: High, scales used in past brand extension research (e.g., Loken & John,
Moderate) between-subjects design. Construal level was mea- 1993). Finally, participants completed several demographic
sured using Vallacher and Wegner's (1989) Behavior Identifica- questions and were thanked and dismissed. They were debriefed
tion Form (BIF). Respondents were presented with two alternative in a classroom setting at the close of the data collection.
descriptions for 25 different target behaviors. For example,
“eating” was described in two ways: one describing eating in Results
terms of how the behavior is performed (e.g., chewing and
swallowing) and one describing eating in terms of why the Preliminary analysis
behavior is performed (e.g., getting nutrition). Respondents were Brand extension measures were analyzed in a 2 (Construal
asked to choose the description that they personally believed to be Level: High, Low) × 2 (Extension Fit: High, Moderate) × 2
more appropriate for each pair. An overall score was obtained by (Brand: Nike, New Balance) analysis of variance. There were
adding the number of abstract descriptions selected by a no significant effects involving the brand factor, neither main
respondent across 25 behaviors. A median split was used to effects nor interactions (p's N .10). Thus, we pooled data for
identify two levels of construal. Specifically, individuals scoring Nike and New Balance for the remaining analyses.
14 or above were classified as “high” construal; individuals with
scores 13 or below were classified as “low” construal. Manipulation checks
Brand extension fit was manipulated at two levels: moderate Manipulations of brand extension fit were examined by
and high. In study 1, we chose Nike and New Balance, which performing a 2 (Extension Fit) × 2 (Construal Level) ANOVA on
were very familiar and equally well liked in a pretest with respondent perceptions of brand extension fit. Two items
participants similar to those in the main study (MNike = 5.23 vs. measuring extension fit were summed for this purpose
120 H. Kim, D.R. John / Journal of Consumer Psychology 18 (2008) 116–126

(Cronbach's α = .95). As expected, the analysis revealed a

significant main effect of fit level (F(1,154) = 65.58, p b .001),
with comfort insoles (M = 5.55) perceived as more consistent with
the Nike/New Balance brand than were treadmills (M = 3.49). In
addition, fit perceptions did not differ for low-construal and high-
construal consumers (p N .10), allowing us to rule out the
possibility that differences in perceived fit were responsible for
construal level differences in brand extension evaluations.3
Differences between low- and high-construal groups were
examined to validate the measures used to assign respondents to a
group. Specifically, we examined the types of thoughts
respondents listed when evaluating brand extensions, coding
thoughts as representative of more abstract generalized thinking
about the brand extension (e.g., stereotypical features of the parent
brand) or representative of more concrete contextualized thinking Fig. 1. Study 1: Extension evaluations as a function of construal level and fit
about the brand extension (e.g., buying or usage context for the level.
extension product). Each respondent's thought listing was coded
for the number of abstract and concrete thoughts by two high (M = 4.99) and moderate fit extensions (M = 4.85) did not
independent coders (82% agreement). An overall measure of differ from one another (p N .20).
abstract thinking was computed for each respondent by adding the
number of abstract thoughts and then subtracting the number of Weight of perceived extension fit
concrete thoughts. A 2 (Extension Fit) × 2 (Construal Level) Compared to low-construal consumers, high-construal con-
ANOVA was performed on this measure, revealing a significant sumers were predicted to place more weight on extension fit as
main effect of construal level. As expected, respondents assigned input into making their brand extension evaluations. To test this
to the high-construal group exhibited more abstract thinking prediction, we developed a regression equation for each
during the brand evaluation process than respondents assigned to construal group, with brand extension evaluation as the
the low-construal group (F(1, 152) = 4.69, p b .05). dependent variable, and brand extension fit and parent brand
attitude as the independent variables:
Brand extension evaluations
Prior to our main analysis, a 2 (Extension Fit) × 2 (Construal Group 1 (People with a low-level construal):
Level) ANOVA was performed on parent brand familiarity and
attitudes to detect whether differences in these factors might Brand Extension Evaluation
confound the results for brand extension evaluations. The ¼ b10 þ b11 PerceivedFit þ b12 BrandAttitude þ e ð1Þ
analysis revealed no significant differences for parent brand
Group 2 (People with a high-level construal):
familiarity or parent brand attitudes across experimental
conditions (p's N .10). Brand Extension Evaluation
A 2 (Extension Fit) × 2 (Construal Level) analysis of variance ¼ b20 þ b21 PerceivedFit þ b22 BrandAttitude þ e ð2Þ
was performed on brand extension evaluations, summing the
three evaluation measures from the survey (α = .85). The results Following the procedure suggested by Gujarati (1970), we
revealed a significant main effect of extension fit level (F(1, compared β11 (i.e., the impact of perceived fit on brand extension
154) = 13.34, p b .001), which was qualified by an expected evaluations among people with low-level construals) and β21
construal level x extension fit interaction (F(1, 154) = 7.16, (i.e., the impact of perceived fit on brand extension evaluations
p b .01; see Fig. 1).4 Consistent with our hypothesis, high- among people with high-level construals). In order to test for
construal participants evaluated the high fit extension (M = 5.50) statistically significant differences across groups, we combined
more favorably than the moderate fit extension (M = 4.56) (F(1, the two equations above into one using dummy coding:
154) = 19.33, p b .001). In contrast, low-construal participants
were not as sensitive to the level of perceived fit in rendering Brand Extension Evaluation ¼ b0 þ b1  D
their evaluations of the brand extensions; their evaluations of þ b2 PerceivedFit
þ b3 PerceivedFit  D
þ b4 BrandAttitude
3
In other situations, it is possible that construal level could influence
þ b5 BrandAttitude  D þ e ð3Þ
perceived fit by affecting the type of information used for judging fit. In our D ¼ 1 ðif a high  level construalÞ
case, we selected extensions that would be perceived equally in terms of fit
¼ 0 ðif a low  level construalÞ
across construal levels to provide a clean test for our extension evaluation
hypotheses.
4
A separate regression analysis performed with construal level as a In terms of interpretation, we can conclude that the impact
continuous variable confirmed these results. Specifically, the interaction term of perceived extension fit for Group 1 (β2) is not the same as
was still significant (β = .086; p b .01). that for Group 2 (β2 + β3) if β3 is significantly different from
 H. Kim, D.R. John / Journal of Consumer Psychology 18 (2008) 116–126 121

zero. That is, β3 indicates the incremental weight of perceived coherent representations that extract the gist from available
extension fit on extension evaluations for Group 1 versus information when an event is expected to occur in the distant
Group 2.5 versus near future, we expect consumers to place more
As expected, we found β3 to be significantly greater than importance on fit perceptions when evaluating brand extensions
zero (β3 = .151; p b .05). Consistent with our hypothesis, that will be available in the distant future versus the near future.
consumers using high-level construals rely on their perceptions Formally,
of extension fit in evaluating brand extensions to a greater
H3. Consumers evaluating a brand extension that will be
degree than do consumers who tend to use low-level construals.
available in the distant future will place more weight on
Additionally, we found that the influence of brand attitude on
perceived extension fit in their extension evaluations compared
extension evaluations was not statistically significant between
to consumers evaluating a brand extension that will be available
groups (β5 = .05; p N .20). Thus, differences in brand extension
in the near future.
evaluations among high- versus low-construal consumers are
due to the differential weighting of perceived fit, and not to Further, we predict that differences in brand extension
differences in parent brand attitudes. evaluations for high versus moderate fit extensions will be
greater for consumers evaluating a brand extension that will be
Discussion available in the distant future. Following the logic presented in
study 1, consumers who place a greater importance on
Our findings support the idea that construal level influences perceived fit will evaluate high fit extensions more favorably
the importance of extension fit in consumer evaluations of than moderate fit extensions. Because consumers evaluating a
brand extensions. High fit extensions were evaluated more brand extension available in the distant future place more
favorably than moderate fit extensions for high-construal importance on perceived fit, it follows that these consumers will
consumers; evaluations did not differ for low-construal evaluate high versus moderate fit extensions more positively
consumers. Additionally, the weight placed on perceived than will consumers who place less importance on perceived fit
extension fit as input into brand extension evaluations was (i.e., consumers who evaluate an extension available in the near
much greater for high-construal than for low-construal future). Formally,
consumers. These patterns were consistent across different
H4. Differences in brand extension evaluations for high versus
types of parent brands, regardless of whether they were broad
moderate fit extensions will be greater for brand extensions that
brands (Nike) or narrow brands (New Balance). Overall, these
will be available in the distant versus near future.
results add to prior research by identifying construal level as a
moderator of the well-accepted relationship between brand
extension fit and brand extension evaluation. Study 2
In the next study, we replicate our results with a different
brand (MTV) and different brand extensions (high fit: music Our hypotheses were tested in a 2 (Temporal Distance: Near
downloading; moderate fit: travel agency). We do so to rule out Future, Distant Future) × 2 (Extension Fit: High, Moderate)
the possibility that something unique to the Nike/New Balance between-subjects design. Construal level was manipulated in
brands or brand extensions drove the results we obtained in the this study by varying the time when the brand extension would
first study. We also manipulate construal level in the next study. be available for purchase.
In study 1, construal level was measured as chronic individual
difference, which could be correlated with other factors Sample and procedure
affecting extension evaluations. To rule out this possibility,
we manipulate construal level in the next study via temporal A total of ninety undergraduates enrolled in introductory
distance. Drawing on the premise that temporally distant (near) marketing courses participated in the study in order to fulfill a
future events tend to be represented with abstract (concrete) course requirement. These participants were given a survey
concepts (Trope & Liberman, 2003), we varied the temporal asking about their opinions concerning new products. First, the
distance of brand extension introductions. Since people are participants encountered temporal distance manipulations from
more likely to use abstract mental models and to rely on the cover sheet. Those assigned to the near future condition
were informed that the new products would be launched very
5
soon. In contrast, those assigned to the distant future condition
Another way of testing equality between sets of coefficients in two linear were informed that the new products would be launched six to
regressions is the Chow test (Chow, 1960). However, from the results of the
Chow test, we cannot determine which sets of coefficients cause differences
eight months from the present. Next, participants were asked to
between the two models, as the Chow test only indicates whether there is a turn the page and examine a recent (fictitious) press release.
structural change in the relationship between the regressed Y and the regressors This press release briefly described: 1) the company name; 2)
(Greene, 2000; Gujarati, 2003). The comparison of the impact of fit perceptions the type of new product; and 3) when the new product would
is our primary interest, so we adopted a dummy coding approach that can be available. An MTV music downloading service and an
pinpoint what causes the structural change. That is, we can examine whether
the impact of perceived fit differs across people with high- versus low-level MTV travel agency were chosen as the high and moderate fit
construals. Similar findings are obtained, however, using a Chow test (F(2, extensions, respectively. After viewing the press release,
152) = 2.94; p b .06). participants were asked for their evaluations of the new
122 H. Kim, D.R. John / Journal of Consumer Psychology 18 (2008) 116–126

MTV services and their perceptions of how these services fit Weight of perceived extension fit
with the MTV brand. Finally, participants were debriefed and We used the analytical procedure from study 1 to test the
thanked. relative weight of fit perceptions between near and distant future
times (see Equation (3)). As expected, we found the key
Results regression coefficient (β3) to be positive and significant
(β3 = .22; p b .05), indicating that participants in the distant
Manipulation check future condition (who favor high-level construals) relied more
Measures of perceived extension fit were analyzed to check on their perceptions of extension fit in evaluating brand
the adequacy of the brand extensions chosen for the study. As extensions than did participants in the near future condition
expected, a 2 (Temporal Distance) × 2 (Extension Fit) ANOVA (who favor low-level construals). Additionally, the influence of
performed on the perceived fit measure (Cronbach's α = .95) brand attitude on extension evaluations was not statistically
revealed a significant main effect of fit level (F(1, 86) = 66.79, significant between groups (β5 = −.20; p N .10). Thus, differ-
p b .001), indicating that the music downloading service was ences in brand extension evaluations in near versus distant
perceived as more consistent with the MTV brand (M = 5.97) future times are due to the differential weighting of perceived
than the travel agency (M = 3.83). Also, fit perceptions did not fit, and not to differences in parent brand attitudes.
differ across temporal distance conditions, allowing us to rule
out the possibility that differences in perceived fit could be Discussion
responsible for differences in brand extension evaluations in the
near versus distant future. Our findings confirm that construal levels moderate the
importance placed on perceived fit in brand extensions.
Brand extension evaluations Consumers who evaluated brand extensions described as
A 2 (Temporal Distance) × 2 (Extension Fit) analysis of available in the distant future (evoking high-level construals)
variance with age as a covariate (ANCOVA) was conducted on placed greater weight on perceived fit in rendering their
extension evaluations, measured by summing four evaluation extension evaluations than did consumers evaluating brand
items on the survey (Cronbach's α = .90). The analysis revealed extensions described as available in the near future (evoking
significant main effects of temporal distance (F(1, 85) = 11.62, low-level construals). Further, differences in extension evalua-
p b .01) and extension fit (F(1, 85) = 20.00, p b .001). More tions for high versus moderate fit extensions were greater for
importantly, these results were qualified by a significant two- consumers judging extensions becoming available in the distant
way interaction between temporal distance and extension fit (F versus near future. Thus, consumers who favor high-level
(1, 85) = 4.16, p b .05; see Fig. 2). Consistent with our construals, based on temporal distance (study 2) or chronic
hypothesis, participants assigned to the distant future condition individual tendencies (study 1), consider perceived fit as more
(with a brand extension launch expected in six to eight months) important input into brand extension evaluations. These
evaluated a high fit extension (M = 5.86) more favorably than a findings are consistent with the idea that construal levels
moderate fit extension (M = 4.43) (F(1, 85) = 20.53, p b .001). In influence extension evaluations by shifting preferences for
contrast, the participants assigned to the near future condition information that matches an individual's abstract or concrete
(with a brand extension launch expected in a week) were not as mindset.
sensitive to the level of perceived fit in rendering their Given these findings, is it possible to increase the importance
evaluations of the extensions (high fit: M = 4.63; moderate fit: placed on perceived fit among consumers who tend to favor
M = 4.09; p N .05). low-level construals? Brand extensions that fit well with the
parent brand usually have a natural advantage in terms of
consumer acceptance, requiring less investment in marketing
communications and sales promotion. But, as we have seen,
low-construal consumers are not swayed by perceived fit,
thereby reducing one of the main factors firms can rely on for
launching successful brand extensions.
In the next study, we examine a strategy for encouraging
low-construal consumers to use perceived fit as important input
into their brand extension evaluations. Recall our earlier
discussion of why high-construal individuals use perceived
extension fit when evaluating brand extensions. Perceptions of
extension fit, which are abstract generalized assessments of the
connections existing between an extension and parent brand,
match the mindsets of consumers that construe stimuli at a more
abstract level (high-level construals). Information matching an
individual's mindset is thought to be more influential because it
Fig. 2. Study 2: Extension evaluations as a function of temporal distance and fit is perceived as more important, more valid, and worthy of more
level. consideration (Lavine & Snyder, 1996; Lee et al., 2000; Petty &
 H. Kim, D.R. John / Journal of Consumer Psychology 18 (2008) 116–126 123

Wegener, 1998). Following this line of reasoning, we provided appendix A). A pretest confirmed that individuals reading this
consumers with information about the importance of fit in article considered fit to be more important and useful in these
several contexts (e.g., marriages, friendships, new products) contexts than did individuals who did not read the article
prior to being exposed to the brand extensions. For low- (M = 6.00 and M = 5.08, respectively; F(1, 27) = 10.35, p b .01).
construal consumers, we expected this information to increase
the importance of perceived fit when evaluating brand Results
extensions, resulting in high fit extensions being evaluated
more favorably than moderate fit extensions. For high-construal Manipulation checks
consumers, we expected the information to have little or no Respondents' perceptions of brand extension fit were
effect, given that perceived fit is already important to these analyzed to check the adequacy of the brand extensions chosen
consumers as evidenced by their more favorable evaluations for for the study. As expected, a 2 (Construal Level) × 2 (Extension
high fit than moderate fit extensions. Thus, we predicted: Fit) × 2 (Information Cue) ANOVA performed on perceived
extension fit revealed a significant main effect of fit level (F(1,
H5a. Without an information cue, differences in brand
129) = 71.98, p b .001), with shoe insoles perceived as more
extension evaluations for high versus moderate fit extensions
consistent with the New Balance brand (M = 5.66) than were
will be greater for consumers with a chronic tendency toward
treadmills (M = 3.44). Thus, the manipulation of fit level was
high-level (vs. low-level) construals.
successful.
H5b. With an information cue, differences in brand extension
evaluations for high versus moderate fit extensions will be Brand extension evaluations
similar for consumers with a chronic tendency toward high- A 2 (Construal Level) × 2 (Extension Fit) × 2 (Information
level and low-level construals. Cue) ANOVA was performed on brand extension evaluations,
averaging the three evaluation measures from the survey
Study 3 (α = .84). The results revealed a significant main effect of
extension fit level (F(1, 129) = 19.77, p b .001) and a significant
A 2 (Construal Level: High, Low) × 2 (Extension Fit: High, main effect of information cue (F(1, 129) = 10.69, p b .01),
Moderate) × 2 (Information Cue: Present, Absent) between- which were qualified by an expected three-way interaction (F(1,
subjects design was used to investigate these predictions. As 129) = 4.51, p b .05; see Fig. 3).6
before, construal level was measured using the Behavior Further analysis of the three-way interaction revealed that the
Identification Form (BIF). Individuals scoring 14 or above two-way interaction between construal level and extension fit was
were classified as “high” construal, whereas individuals with significant when the information cue was absent (F(1, 129) = 4.98,
scores of 13 or below were classified as “low” construal. New p b .05), replicating our previous findings. Without an information
Balance was the focal brand, and the same types of brand cue, high-construal consumers evaluated the high fit extension
extensions used in study 1 (shoe insoles as a high-fit extension (M = 5.67) more favorably than the moderate fit extension
and treadmills as a moderate-fit extension) were employed in (M = 4.53) (F(1, 129)= 16.58, p b .001); low-construal consumers
this study. An information cue, consisting of an article about fit did not differ in their evaluations of the high fit (M = 5.07) and
being important in many different contexts, was developed to moderate fit (M = 4.88) extensions (p N .35). However, when an
increase perceptions of perceived fit as being important and information cue was provided, the two-way interaction between
useful for brand extension evaluations. extension fit and construal level was not significant (F b 1). Of
interest here is the fact that low-construal consumers evaluated the
Sample and procedure high fit extension (M = 4.95) more favorably than the moderate fit
(M = 3.89) extension (F(1, 129)= 9.50, p b .01), which is a pattern
One-hundred-thirty-seven undergraduates participated in the typically found only for high-construal consumers. Thus, an
study in order to fulfill a course requirement. The experimenter information cue encouraged low-construal consumers to use
informed participants that they would be asked to complete a perceived fit in rendering their brand extension evaluations, as
number of unrelated tasks during the study. Instructions and expected.
questions for each task were included in a survey booklet, with
the section for each task separated by a distinct heading and Discussion
typed in a different font and font size, which furthered
perceptions that the tasks were unrelated to each other. We find that low-construal consumers can be encouraged to
Participants were first asked about their attitudes toward several use perceived fit in evaluating brand extensions by cueing them
brands. Next, one-half of the respondents were asked to perform to the importance of perceived fit. Without an information cue,
a task prior to the main brand extension evaluation task. these consumers evaluated high and moderate fit extensions
Participants in this condition were asked to read an article similarly. With an information cue, low-construal consumers
entitled “Successful Marriages, Friendships, and New Products:
Fit is Important!” and evaluate the article on several dimensions 6
A separate regression analysis performed with construal level as a
related to writing quality. The article described fit as a good continuous variable confirmed these results. Specifically, the three-way
predictor of relationship success and new product success (see interaction term was marginally significant (β = −.12; p b .07).
124 H. Kim, D.R. John / Journal of Consumer Psychology 18 (2008) 116–126

been identified as the primary determinant for the success of

brand extensions (Bottomley & Holden, 2001; Völckner &
Sattler, 2006). Research has neglected the possibility that
perceived fit may be more or less influential depending on
factors that exist at the individual or situational level.
We find that construal level is a moderator of the relationship
between perceived fit and brand extension evaluations. Results
from three studies indicate that consumers who construe stimuli
at a higher level, based on a chronic tendency (studies 1 and 3)
or a situational factor (study 2), place more importance on
perceived fit in evaluating brand extensions. These consumers
are more sensitive to the level of perceived fit, evaluating high
fit extensions more favorably than moderate fit ones. Con-
sumers who construe their environments at a lower level are less
sensitive to the level of perceived fit, evaluating moderate
versus high fit extensions in a similar way.

Theoretical contributions

Our findings contribute to an emerging stream of research

that explores personal, situational, and cultural differences in
consumer responses to brand extensions. To date, researchers
have identified several factors that affect how consumers judge
perceived fit, including cultural differences in styles of thinking
(Monga & John, 2007; Ng & Houston, 2006) and situational
differences in mood (Barone & Miniard, 2002; Barone et al.,
2000). For example, researchers have found that consumers
from Eastern cultures, with a more holistic style of thinking,
perceive higher levels of extension fit than do consumers from
Western cultures, with a more analytic style of thinking (Monga
& John, 2007; Ng & Houston, 2006). Our findings suggest that
Fig. 3. Study 3: extension evaluations as a function of construal level, fit level
personal, situational, and cultural factors may also influence the
and information cue.
weight that consumers place on fit perceptions once these
perceptions have been formed. Once consumers have found a
evaluated high fit extensions more favorably than moderate fit link connecting a brand extension to the parent brand, factors
extensions, consistent with evaluations of high-construal such as construal level determine whether or not perceptions of
consumers. fit are used in evaluating brand extensions.
Thus, by increasing perceptions of the importance of Why do some consumers place more weight on perceived fit
extension fit, low-construal consumers responded to brand in the process of evaluating brand extensions? We have
extensions as if they were high-construal consumers. This reasoned that perceived fit is an abstract generalized assessment
finding suggests that communication strategies may be able to of the connections that exist between a brand extension and the
overcome the tendency of low-construal consumers to ignore parent brand. As such, perceived fit is more likely to be used by
extension fit. Moreover, our findings provide further evidence consumers who construe their environments in terms of
that differences in brand extension evaluation for low- versus abstract generalized features (high-level construals). Thus,
high-construal consumers are due to the importance that construal level influences extension evaluations by making
perceived fit plays in the evaluation process. High-construal perceived fit more important when it matches an individual's
consumers use perceived fit as a diagnostic cue for evaluating abstract mindset. This view is consistent with our findings from
brand extensions, but low-construal consumers only do so when study 3, where we manipulated the importance of fit
they receive information cueing them to the importance of perceptions using an information cue. When fit perceptions
perceived fit. were described as important and useful in judgment contexts,
low-construal consumers evaluated brand extensions on the
General discussion basis of extension fit. When no cue was given, low-construal
consumers reverted to their natural inclinations to ignore
One of the most well established findings from brand perceived fit in evaluating extensions. Consistent with the
extension research is that consumers evaluate brand extensions accessibility-diagnosticity perspective (Ahluwalia & Gürhan-
on the basis of their perceived fit with the parent brand. In fact, Canli, 2000; Feldman & Lynch, 1988), our research suggests
the degree to which an extension fits with a parent brand has that extension fit is deemed as more diagnostic to consumers
 H. Kim, D.R. John / Journal of Consumer Psychology 18 (2008) 116–126 125

who have a tendency to construe their environments in more and temporal distance. However, sensory distance (Kardes et
abstract generalized terms. al., 2006), spatial distance (Fujita et al., 2006), and power
priming (Smith & Trope, 2006) have also been found to
Directions for future research influence mental construal levels. These dimensions suggest
additional avenues for future research examining the role of
Our findings raise a number of issues worthy of future construal level on consumer response to brand extensions.
research. First, we might consider the possibility that construal
level influences judgments of perceived fit, in addition to the Acknowledgments
construal level effects on the weight placed on perceived fit in
evaluating brand extensions. As noted in Lynch and Zauberman The authors thank Jae Eun Kim, Ji Kyung Park, and Sara
(2007), psychological distance can influence both mental John for their assistance in data collection and analysis.
representation of inputs and the effective weights given to
decision criteria. In our studies, we selected brand extensions Appendix A. Article used in the information cue condition
that evoked the same fit perceptions among high- and low-
construal consumers to provide a cleaner test of our predictions. Successful marriages, friendships, and new products: fit is
However, the description of consumers with a tendency toward important!!
low- versus high-level construals suggests that these groups
could have different perceptions of brand extension fit as well, Ernest W. Burgess, a sociology professor at Harvard
based on the type of information (abstract, concrete) they use for University, has spent the last twenty years conducting research
thinking about extension fit. Thus, although this research on successful marriages. While a number of factors are
illuminates a new area where construal levels can affect important, Dr. Burgess has found that marital success depends
consumers' judgment and decision making, there remains primarily on the “fit” between marriage partners—that is, how
room for research on the conditions under which people with similar they are in terms of basic values, interests, and hopes.
different construal levels generate different fit perceptions. Similarly, researchers have found that long-lasting friend-
Second, it may be useful to examine the role that construal ships, like marriages, also depend on a good fit between two
level plays in designing successful marketing communications for people. Irrespective of race, gender, and age, the better two
brand extensions. Marketers could decide whether or not to people's needs, attitudes, personalities, and interests “fit”
highlight the fit between a new product and the parent brand, together, the more rewarding and long lasting the relationships
given that fit is more or less important depending on construal between them will be.
level. For high-construal consumers, we would expect that As such, the quality of match (or fit) is an important factor
messages communicating how well the extension fits with the influencing the success of relationship among human beings.
parent brand, along with other abstract or generalized information However, the importance of perfect fit can be also seen in
(e.g., quality awards, the Good Housekeeping seal), would be business practices, such as new product development. In fact,
more persuasive than messages describing specific attributes and both business analysts and strategic consultants strongly argue
usage contexts. In contrast, for low-construal consumers, we that fit is all about successful new product launching.
would expect messages focusing on more specific and contextual A recent study by McKinsey and Company proves that new
information, such as detailed information (e.g., length of products that are similar to other products produced by a firm are
warranty) or attribute-aligned comparisons to other brands more successful. Firms such as 3 M and P&G have had many new
(Malkoc, Zauberman, & Ulu, 2005), would be more persuasive. product successes closely tied to products they already make.
Third, future research might examine the moderating effect of Other firms, such as Eddie Bauer and Starbucks, have been
construal level for different types of brand extensions. In our successful launching new products that consumers can use with
studies, we examined high and moderate fit extensions, but not their existing products. For example, Starbucks has launched a
low fit extensions. Although managerial interest is usually successful line of cappuccino makers that consumers can use with
focused on the high and moderate fit levels, low fit extensions the ground coffee purchased at their neighborhood Starbucks.
might produce additional insights into the differences between In sum, researchers in different areas all agree that “fit” is
low- and high-construal consumers. Two outcomes seem important—whether you are looking for a successful marriage,
possible. One, consistent with our reasoning, the low fit extension long-lasting friendship, or great new product idea!
might be evaluated less favorably by high-construal consumers
but not low-construal consumers. Two, the fit between the References
extension and parent brand may be so obviously poor that both
low-construal and high-construal consumers may use perceived Aaker, D. A., & Keller, K. L. (1990). Consumer evaluations of brand extensions.
fit to arrive at an unfavorable extension evaluation. Journal of Marketing, 54, 27−41.
Finally, one could examine additional factors that influence Aaker, J. L., & Lee, A. Y. (2001). “I” seek pleasures and “we” avoid pains: The
construal levels since the implications of construal level role of self-regulatory goals in information processing and persuasion.
Journal of Consumer Research, 28, 33−49.
facilitate a better understanding of preference formation (Kardes Ahluwalia, R., & Gürhan-Canli, Z. (2000). The effects of extensions on the
et al., 2006; Liberman & Trope, 1998). In our studies, we family brand name: An accessibility-diagnosticity perspective. Journal of
examine construal level as a function of individual differences Consumer Research, 27, 371−381.
126 H. Kim, D.R. John / Journal of Consumer Psychology 18 (2008) 116–126

Ashmore, R. D., & Del Boca, F. K. (1981). Conceptual approaches to Keller, K. L. (2002). Branding and brand equity. Cambridge, MA: Marketing
stereotypes and stereotyping. In D. L. Hamilton (Ed.), Cognitive processes Science Institute.
in stereotyping and intergroup behavior (pp. 1−35). Hillsdale, NJ: Keller, K. L., & Aaker, D. A. (1992). The effects of sequential introduction of
Erlbaum. brand extensions. Journal of Marketing Research, 29, 35−50.
Barone, M. J., & Miniard, P. W. (2002). Mood and brand extension judgments: Lane, V. R. (2000). The impact of ad repetition and ad content on consumer
Asymmetric effects for desirable versus undesirable brands. Journal of perceptions of incongruent extensions. Journal of Marketing, 64, 80−91.
Consumer Psychology, 12, 283−290. Lavine, H., & Snyder, M. (1996). Cognitive processing and the functional
Barone, M. J., Miniard, P. W., & Romeo, J. B. (2000). The influence of positive matching effect in persuasion: The mediating role of subjective perceptions
mood on brand extension evaluations. Journal of Consumer Research, 26, of message quality. Journal of Experimental Social Psychology, 32,
386−400. 580−604.
Bottomley, P. A., & Holden, S. J. S. (2001). Do we really know how consumers Lee, A. Y., Aaker, J. L., & Gardner, W. L. (2000). The pleasures and pains of
evaluate brand extensions? Empirical generalizations based on secondary distinct self-construals: The role of interdependence in regulatory focus.
analysis of eight studies. Journal of Marketing Research, 38, 494−500. Journal of Personality and Social Psychology, 78, 1122−1134.
Boush, D. M., & Loken, B. (1991). A process-tracing study of brand extension Liberman, N., & Trope, Y. (1998). The role of feasibility and desirability
evaluation. Journal of Marketing Research, 28, 16−28. considerations in near and distant future decisions: A test of temporal
Boush, D. M., Shipp, S., Loken, B., Gencturk, E., Crockett, S., Kennedy, E., et al. construal theory. Journal of Personality and Social Psychology, 75, 5−18.
(1987). Affect generalization to similar and dissimilar brand extensions. Liberman, N., Trope, Y., & Wakslak, C. (2007). Construal level theory and
Psychology & Marketing, 225−237. consumer behavior. Journal of Consumer Psychology, 17, 113−117.
Bridges, S., Keller, K. L., & Sood, S. (2000). Explanatory links and the perceived Loken, B., & John, D. R. (1993). Diluting brand beliefs: When do brand
fit of brand extensions: The role of dominant parent brand associations and extensions have a negative impact? Journal of Marketing, 57, 71−84.
communication strategies. Journal of Advertising, 29, 1−11. Lynch, J. G., Jr., & Zauberman, G. (2007). Construing consumer decision
Broniarczyk, S. M., & Alba, J. W. (1994). The importance of the brand in brand making. Journal of Consumer Psychology, 17, 107−112.
extension. Journal of Consumer Research, 31, 214−228. Malkoc, S. A., Zauberman, G., & Ulu, C. (2005). Consuming now or later? The
Chow, G. (1960). Tests of equality between sets of coefficients in two linear interactive effect of timing and attribute alignability. Psychological Science,
regressions. Econometrica, 28, 591−605. 16, 411−417.
Dhar, R., & Kim, E. Y. (2007). Seeing the forest or the trees: Implications of Monga, A. B., & John, D. R. (2007). Cultural differences in brand extension
construal level theory for consumer choice. Journal of Consumer evaluation: The influence of analytic versus holistic thinking. Journal of
Psychology, 17, 96−100. Consumer Research, 33, 529−536.
Edwards, K. (1990). The interplay of affect and cognition in attitude formation Ng, S., & Houston, M. J. (2006). Exemplars or beliefs? The impact of self-view
and change. Journal of Personality and Social Psychology, 59, 202−216. on the nature and relative influence of brand associations. Journal of
Fabrigar, L. R., & Petty, R. E. (1999). The role of the affective and cognitive Consumer Research, 32, 519−529.
bases of attitudes in susceptibility to affectively and cognitively based Nussbaum, S., Trope, Y., & Liberman, N. (2003). Creeping dispositionism: The
persuasion. Personality and Social Psychology Bulletin, 25, 363−381. temporal dynamics of behavior prediction. Journal of Personality and
Feldman, J. M., & Lynch, J. G. (1988). Self-generated validity and other effects Social Psychology, 84, 485−497.
of measurement on belief, attitude, intention, and behavior. Journal of Nussbaum, S., Liberman, N., & Trope, Y. (2006). Predicting the near and distant
Applied Psychology, 73, 421−435. future. Journal of Experimental Psychology: General, 135, 152−161.
Fiske, S. T., & Pavelchak, M. A. (1986). Category-based versus piecemeal-based Park, C. W., Milberg, S., & Lawson, R. (1991). Evaluation of brand extensions:
affective responses: Developments in schema-triggered affect. In R. M. The role of product feature similarity and brand concept consistency.
Sorrentino, & E. T. Higgins (Eds.), Handbook of motivation and cognition: Journal of Consumer Research, 18, 185−193.
Foundations of social behavior (Vol. 1, pp. 167–203). New York: Guilford Petty, R. E., & Wegener, D. T. (1998). Matching versus mismatching attitude
Press. functions: Implications for scrutiny of persuasive messages. Personality and
Freitas, A. L., Salovey, P., & Liberman, N. (2001). Abstract and concrete self- Social Psychology Bulletin, 24, 227−240.
evaluative goals. Journal of Personality and Social Psychology, 80, 410−424. Smith, P. K., & Trope, Y. (2006). You focus on the forest when you're in charge
Fujita, K., Henderson, M. D., Eng, J., Trope, Y., & Liberman, N. (2006). Spatial of the trees: Power priming and abstract information processing. Journal of
distance and mental construal of social events. Psychological Science, 17, Personality & Social Psychology, 90, 578−596.
278−282. Trope, Y., & Liberman, N. (2000). Temporal construal and time-dependent
Greene, W. H. (2000). Econometric analysis, 4th ed. Upper Saddle, NJ: Prentice changes in preference. Journal of Personality and Social Psychology, 79,
Hall. 876−889.
Gujarati, D. M. (1970). Use of dummy variables in testing for equality between Trope, Y., & Liberman, N. (2003). Temporal construal. Psychological Review,
sets of coefficients. American Statistician, 24, 50−52. 110, 403−421.
Gujarati, D. N. (2003). Basic econometrics (4th ed.). New York: McGraw Hill. Trope, Y., Liberman, N., & Wakslak, C. (2007). Construal levels and
Han, J. K., & Schmitt, B. H. (1997). Product-category dynamics and corporate psychological distance: Effects on representation, prediction, evaluation,
identity in brand extensions: A comparison of Hong Kong and U.S. and behavior. Journal of Consumer Psychology, 17, 83−95.
consumers. Journal of International Marketing, 5, 77−92. Vallacher, R. R., & Wegner, D. M. (1987). What do people think they're
Herr, P. M., Farquar, P. H., & Fazio, R. H. (1996). Impact of dominance and doing? Action identification and human behavior. Psychological Review,
relatedness on brand extensions. Journal of Consumer Psychology, 5, 135−159. 94, 3−15.
Hilton, J. L., & von Hippel, W. (1996). Stereotypes. Annual Review of Vallacher, R. R., & Wegner, D. M. (1989). Levels of personal agency: Individual
Psychology, 47, 237−271. variation in action identification. Journal of Personality and Social
Kardes, F. R., Cronley, M. L., & Kim, J. (2006). Construal-level effects on Psychology, 57, 660−671.
preference stability, preference-behavior correspondence, and the suppres- Völckner, F., & Sattler, H. (2006). Drivers of brand extension success. Journal
sion of competing brands. Journal of Consumer Psychology, 16, 135−144. of Marketing, 70, 18−34.